David Peterson wrote:

> Speaking of which, I had an idea that I haven't fully fleshed out
> yet (too many variables) about a single, unified romanization
> system that we could use for everything. The reason is that unless
> one's language is spoken by fictional speakers that have access
> to the Roman alphabet, a romanization is not an orthography.
> As such, it's nothing more than a tool, and it'd be nice if there
> was a single way to apply it to every language.

I went through a period where I was trying to use a unified romanization 
for Kolagian languages. Previously each language had its own system of 
romanization, often corresponding with the elements of the native 
writing system. I used probably close to all the diacritics in the 
Latin-1 character set for the vowels, and digraphs for consonants 
(mainly digraphs with -h, but also things like "ng"). Since Latin-1 has 
very few diacritics for consonants, I ended up using things like 
superscript 1, 2, and 3 (¹ ² ³). (I think those were for palatalized, 
velarized, and retroflex although I don't recall specifically.)

At one time I tried coming up with a unified system for Azirian 
languages, using things like the "comma below" diacritic for palatal 
sounds, circumflex for retroflex, prefixing the letter "c" for clicks, 
etc, but I never got very far with that, and currently each language has 
its own romanization. I've considered just using IPA, and in fact my 
current Zharranh romanization uses the IPA symbols for retroflex consonants.

> How it gets implemented, though, depends on what phonemes
> are present in the language. So, for example, if there are aspirated
> stops, use "th", "ph" and "kh" for them. If there are no aspirated
> stops but some (contrastive) fricatives in the same places of
> articulation, though, use these digraphs for those (so "th" for
> voiceless interdental; "kh" for voiceless velar; "ph" for voiceless
> bilabial if voiceless labio-dental is already present). If there are
> both aspirated stops and fricatives like these, then perhaps one
> would go to Unicode.

Well, if you've got a language with unusual sounds like a retroflex 
lateral fricative (as Lindiga has), any way you write that is likely to 
be a bit ad-hoc unless you have a systematic way of representing either 
retroflex sounds, or lateral fricatives, or both. So unless the 
romanization scheme is inclusive enough to account for all the sounds of 
language, even potentially unanticipated sounds, there will be cases 
where it won't fit (in this case, even IPA doesn't have a representation 
for the sound in question, although we can fake it and write [ɬ̢]).

> But the idea is that this would be a program which would allow
> you to feed it a phonology, and it would return a romanization
> system. The idea would be to regularize non-orthographic
> romanization systems, and to cut down on the peculiarities
> (randomly using "x" for [T] or "jh" for [p], or whatever other strange
> thing someone has come up with). It would certainly make
> perusing conlang websites a lot simpler. A romanization system
> shouldn't require a manual--that's what orthographies are for!

Sounds like a worthy goal. Going back to Zharranh, it has a pecular 
letter romanized as "q" that can be pronounced [kw] or [p] depending on 
context. Now years later, I don't remember the exact phonetic rule, so 
it would have been much simpler if I'd just written "kw" or "p" in the 
first place.